Real-time payment fraud detection API — XGBoost (AUC-ROC 0.927) + FastAPI + Docker | Trained on IEEE-CIS dataset (590K transactions)
-
Updated
May 12, 2026 - Jupyter Notebook
#
ieee-cis
Here are 4 public repositories matching this topic...
ML model to predict the probability of fraudulent online transactions — based on the IEEE-CIS Fraud Detection Kaggle competition using real-world e-commerce data from Vesta Corporation.
python data-science machine-learning random-forest scikit-learn kaggle feature-engineering fraud-detection ieee-cis-fraud-detection xgboost-classification ieee-cis
-
Updated
Mar 27, 2026 - Jupyter Notebook
XGBoost fraud detection pipeline with advanced feature engineering for IEEE-CIS, focusing on cardholder UID, D‑column normalization, and group aggregations.
python data-science machine-learning numpy scikit-learn pandas credit-card-fraud kaggle xgboost classification feature-engineering normalization fraud-detection gradient-boosting frequency-encoding ieee-cis group-aggregation
-
Updated
Mar 21, 2026 - Python
A production-ready fraud detection system using Random Forest & SMOTE. Handles severe class imbalance (3.5% fraud rate) and 80%+ missing data. Achieves 0.31 F1-score on IEEE-CIS dataset because of RAM Limitation for applying SMOTE and RF training requiring downsampling of Data
data-science machine-learning random-forest credit-card-fraud kaggle smote fraud-detection imbalanced-classification ieee-cis
-
Updated
May 17, 2026
Improve this page
Add a description, image, and links to the ieee-cis topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the ieee-cis topic, visit your repo's landing page and select "manage topics."